 |
| Titel |
Using conditional probability method for estimation of landslide occurrence probability |
| VerfasserIn |
Meei-Ling Lin, Sheng-Chi Lin |
| Konferenz |
EGU General Assembly 2011
|
| Medientyp |
Artikel
|
| Sprache |
Englisch
|
| Digitales Dokument |
PDF |
| Erschienen |
In: GRA - Volume 13 (2011) |
| Datensatznummer |
250054505
|
|
|
|
|
|
| Zusammenfassung |
| Landslide risk assessment consists of occurrence probability for spatial and temporal
distributions and the concurrent scenario for damage and loss estimation of elements at the
risk. Although items of the landslide risk elements and assessment processes have been
proposed by previous researchers, process for estimating the occurrence probability of
landslide disaster and risk assessment remains a difficult issue. In this study, we adopt the
conditional probability to estimate the occurrence probability of landslide. We propose using
likelihood ratio function method (Chung and Fabbri, 2003) to evaluate spatial probability in
neighboring region, which deals with the insufficient historic landslide data for local
area and avoid unreasonable results caused by extrapolation in statistics method.
The likelihood ratio function for the spatial probability is based on the conditional
probability derived from Bayes’s formula, and the annual probability of occurrence is
estimated based on the probability of landslide events triggered in the local area
given the typhoon occurrences. Therefore, the joint probability of occurrence can
be obtained by combining the spatial probability and temporal probability. The
resulting conditional probability method has been applied to case study in Taiwan; in
particular, the Lushan study area with field monitoring system (Lin, et al. 2007).
Through evaluation of likelihood ratio method, the resident area is within the spatial
probability greater than 80%, and the annual probability of landslide induced by
typhoons is 22% according to the historic event data. Thus the overall annual landslide
occurrence probability for resident area is greater than 18%, and the accumulated
occurrence probability is larger than 50% over the five-year period. According
to the field monitoring data since 2007 including the displacement records from
extensometer, the ground-based Lidar, and close-range photogrammetry surveys
all provided evidences of sliding, and the observed ?eld sliding conditions were
considered consistent with the results of landslide occurrence probability analysis. |
|
|
|
|
|
|